Diagnostic value of four coagulation function indicators for distinguishing between tuberculous pleural effusions and malignant pleural effusions due to lung adenocarcinoma

Background: Coagulation function indicators are often used to evaluate patients’ coagulation function.They are rarely used in disease diagnosis;their use is limited to somecoagulopathies. We described the pattern of coagulation function indicators in patients with pleural effusions and determined the diagnostic value of different abnormal patterns of these indicators for distinguishing tuberculous pleural effusions (TPEs) from malignant pleural effusions associated with lung adenocarcinoma (AD-MPEs). Methods: Patients with AD-MPEs or TPEs participated in this retrospective study.Coagulation functionindicators, including activated partial thromboplastin time, prothrombin time (PT), brinogen (FIB), thrombin time, brinogen degradation products (FDP), and D-dimer(D-D), were measured; their levels and abnormal patterns were analyzed separately and in combination,to determine the diagnostic performance of individual and combined values and the optimal diagnostic model. Results: One hundred patients with AD-MPEs and 84 patients with TPEs participated.PT, FIB, FDP, and D-Dlevels in patients with TPE were signicantly higher than those in patients with AD-MPE (all P<0.01).Receiver operating characteristiccurves revealed that PT, FIB, FDP, and D-Dvalues, particularly FDP and D-D, could distinguish betweenAD-MPE and TPE. The diagnostic performance was better for the combination of the four indicators than for any single indicator (Youden’s index: 0.513; area under the curve: 0.810).The proportion of patients with abnormal patterns of all four indicators was signicantly higher in those with TPE than in those with AD-MPE (P<0.001). Conclusions: Thecombined values of PT, FIB, FDP, and D-Dhave diagnostic value for distinguishing TPE from AD-MPE.

Diagnostic value of four coagulation function indicators for distinguishing between tuberculous pleural effusions and malignant pleural effusions due to lung adenocarcinoma Rui  In China, the main causes of exudative pleural effusions are tuberculosis and lung adenocarcinoma [1].
Patients with lung adenocarcinoma often have pleural effusions, of which about 20% are the rst sign of lung adenocarcinoma, and 30% to 40% develop pleural effusions in the course of the disease, which are indicative of a poor prognosis and short survival [2]. Tuberculous pleurisy is the main cause of pleural effusions in areas with a high prevalence of tuberculosis [3]. Although there are some clinical similarities between pleural effusions associated with lung adenocarcinoma (AD-MPE) and tuberculous pleural effusions (TPE), the diagnosis, treatment and prognosis of the two types of effusion are quite different.
The similarities between AD-MPE and TPE,their delayed onset, and invasiveness of the pathology pose a major challenge in the differential diagnosis of these two conditions [4].
In recent years, some novel markers have been reported for the differential diagnosis of pleural effusions, but few have been successfully applied in clinical practice [5][6][7]. Coagulation function can be assessed using a very basic test and is often tested in patients admitted to hospital.Coagulation function testing is often used to evaluate the coagulation function of patients undergoing surgery and for the monitoring of anticoagulant therapy [8][9][10][11]. It is rarely used in disease diagnosis, and when it is, its use is limited to diagnosing certain coagulo pathies [12]. To date, there have been no studies of the levels and the patterns of coagulation function indicators for distinguishing patients with TPE from AD-MPE. The objectives of this study were to describe the pattern of coagulation function indicators in patients with pleural effusions, and to determine the diagnostic value of different abnormal patterns of these indicators for distinguishing TPE from AD-MPE.

Patients
We conducted a retrospective analysis of data from patients diagnosed with pleural effusions at the Tangdu Hospital, Fourth Military Medical University from September 2017 to December 2018. All patients had a diagnostic work-up and the cause of the PE was based on a combination of clinical ndings, histopathology, cytology, or empirical diagnosis(response to therapy).
In order to be eligible for inclusion in the analysis, all patients had to satisfy the following conditions: 1) no bleeding tendency, hematopoietic dysfunction, or taking anticoagulant drugs, no anti-tumor or antituberculosis treatment prior to admission; 2) The clinical data were complete; 3) have no obvious abnormalities in heart, liver or kidney function; 4) no other tumors apart from lung adenocarcinoma.
This study was approved by the Ethics Committee of Tangdu Hospital, Fourth Military Medical University (TDLL-201709-24), and informed consent was waived on the grounds that it was a retrospective study.

Collection and detection of plasma samples
Prior to any cancer and tuberculosis-directed therapy, a sample of 2 ml fasting venous bloodwascollected from each patient in a 2 ml tube containing 0.109 mol/L sodium citrate at a ratio of one-part citrate to nine parts of blood, and centrifuged for 10 minutes at 2500 g.
Indicators of coagulation function were measured using the turbidimetry assay (SYSMEX, Kobe, Japan). Indicators that differed between the two groups were analysed further using receiver operating characteristic (ROC) curves. ROC curves were plotted for each coagulation indicator, and for the indicators combined to determine the optimal cut-point values. Values of each coagulation indicator were classi ed as either normal or elevated based on the cut-points determined using the ROC analysis.The optimal cut-off value was determined at the maximum value for Youden's index (YI). The categorical values of the coagulation indicators in the patients with AD-MPE and TPE were compared using chi square tests. P-values< 0.05 were regarded as statistically signi cant.

Results
Comparison of basic characteristics and coagulation indicators in patients with pleural effusions due to lung adenocarcinoma and tuberculosis Data from 184 patients, 100 with AD-MPE and 84 with TBE, were used in the analysis. Table 1 and  (Table 3, Figure 3). For PT, FIB, FDP and D-D, using AD-MPE as a reference, the optimal discrimination of TPE was determined at cut-off value of 11.65 s, 4.69 g/L, 0.69 µg/mL and 0.25 µg/mL, respectively.Of the four indicators, FIB has the highest sensitivity and the lowest speci city and YI,while FDP had the highest speci city and the lowest sensitivity; D-D had the highest YI and AUC; and the YI of each of the four indicators were < 0.5.When PT, FIB, FDP and D-D were used in combination and presented by ROC curve, the diagnostic performance was improved (Table 3).

Discussion
Using blood and PE markers is useful for distinguishing TPEs from AD-MPEs, such as adenosine deaminase and carcinoembryonic antigen [13,14]. A number of markers can be used to predict TPEs and AD-MPEs, although their use of markers is considerably less speci c than histology. However, the results can complement each other [15]. As a routine diagnostic work-up, PT, FIB, FDP and D-D are often used to manage coagulation and brinolysis in patients; however, this is the rst study to utilize coagulation indicators to distinguish TPEs from AD-MPEs.
Many studies on coagulation function and tumors have found that coagulation indicators, especially FIB, FDP and D-D, are useful for predicting the risk of occurrence, progression, metastasis, and prognosis of lung cancer: Piccioli [16] found that the probability of thromboembolism in patients with tumors was close to 10%; Aminian et al. [17] found that the hypercoagulable state of tumor patients was closely related to cancer progression and prognosis; Palumbo et al. [18] found that FIB and platelets can increasethe possibility of tumor invasion by impeding natural killer cell-mediated elimination of tumor cells during tumor progression. However, there are few reports on the role of coagulation indicators in individuals with tuberculosis [19,20], and there are even fewer reports on coagulation indicators in relation to TPE [21].Therefore, at the outset of this study, we hypothesized that the severity of coagulation dysfunction in patients with AD-MPE swould be more marked than in those TPEs. However, the result was the opposite of what we expected.This study found that the levels of APTT,PT,FIB,FDP and D-Dwere higher, and abnormal patterns were more prevalent in patients with TPEs than in those withAD-MPEs. The results suggest that the severity of coagulation dysfunction is more marked in individuals with benign TPEs than in those with AD-MPEs.
It has been reported that tissue factors are activated and released during acute infection of tuberculosis [22]. As the initiating factor of the endogenous coagulation pathway, tissue factor plays a very important role. PT is the primary indicator with which to monitor the endogenous coagulation pathway [23].
However,there have been no reports on the application of PT to monitor the changes of coagulation function in patients with TPE. In this study, it was found that >70% of the patients with TPE had a high PT. whereas <50% of patients with AD-MPE had a high PT, and this difference was statistically signi cant. This result differs from the consensus perception that lung cancer, especially lung adenocarcinoma, is more likely to cause coagulation dysfunction [24]. Therefore. in clinical practice, attention should be paid not only to FDP and D-D, but also to PT when acute tuberculosis is suspected.
When Mycobacterium tuberculosis invades the pleura, it activates the monocyte macrophage system and the bacteria are swallowed and degraded by macrophages. The degradation products induce a strong speci c allergic reaction and destroy vascular endothelial cells [25]. Progress of lung adenocarcinoma leads to the destruction and disintegration of the tissue, resulting in non-speci c in ammatory reactions and in ammatory mediators. The continuous in ammatory state constitutes the microenvironment of tumor progress [26]. About 90% of patients with malignant tumors have vascular endothelial cell injury, and this is most marked in people with lung adenocarcinoma [27]. The damage to vascular endothelium caused by pleural effusions can destroy the balance of the coagulation and brinolysis systems. For this reason, this study considered all possible abnormal patterns of coagulation function indicators in the two groups, and ten patterns of abnormal coagulation function caused by AD-MPE and TPE, which not only con rmed the results of previous studies, but also re ected the complexity of the two conditions. In this study, simultaneous abnormalities of PT, FIB, FDP and D-D,and PT, FDP and D-D,were present in 44% and 23%of patients with TPE, which was higher than the 10% of these abnormalities in patients with AD-MPE(P<0.05). These results indicate that clinicians should consider the possibility of TPE in patients with pleural effusions who have either of these two abnormal patterns of coagulation function indicators.
As degradation products of cross-linked brin, FDP and D-D are considered to be general indicators of coagulation activation and brinolysis [28][29][30]. In this study, FDP and D-D were the two indicators with the highest diagnostic performance, and FIB had the highest sensitivity, but their YIs, which are based on the sensitivity and speci city of the test were less than 0.5. Therefore, in order to improve the power of these indicators to distinguish between AD-MPE and TPE, we combined PT, FIB, FDP and D-Dto create a joint indicator.ROC curve analysis showed that the diagnostic performance of the combined indicator was better than that of any of the indicators alone. Therefore, although the determination of coagulation function is the most basic method for the differential diagnosis of AD-MPE and TPE, it had the characteristics of simplicity, economy, rapidity and is non-invasive, and is suitable for use in limitedresource settings and in community-based health facilities as an ideal way to distinguish TPE from AD-MPE.
The primary limitation of this study was that participation was restricted to patients withAD-MPEsor TPEs, and patients with lung cancer and tuberculosis without pleural effusions were excluded.We are planning to conduct a further study with a larger sample size that takes into account the pathological classi cation of lung cancer, incorporates more types of tuberculosis, and reduces bias as far as possible, to further explore the value of using coagulation indicators in different pathological types of lung cancer and different types of tuberculosis.

Conclusions
Coagulation function indicators, especially if considered in combination, had relatively good sensitivity and speci city for distinguishing between pleural effusions due to lung adenocarcinoma and those due to tuberculosis. As measuring coagulation function indicators is a routine test, this method may be applied to the differential diagnosis of pleural effusions in community care settings.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used during the current study available from the corresponding author on reasonable request.